Introducing: ConformalPrediction.jl

Predictive Uncertainty Quantification in Machine Learning

Patrick Altmeyer

Delft University of Technology

July 7, 2023

🥜 CP in a Nutshell

Counterfactual Explanation (CE) explain how inputs into a model need to change for it to produce different outputs.

Provided the changes are realistic and actionable, they can be used for Algorithmic Recourse (AR) to help individuals who face adverse outcomes.

🎉 ConformalPredictions.jl: Overview

🕚 Talk Agenda

  1. Whistle-stop tour of Conformal Prediction (CP)
  2. Examples
    1. Conformal Prediction for LLMs
    2. Conformal Prediction for Explainability (CounterfactualExplanations.jl)
    3. Conformal Bayesian Neural Networks (LaplaceRedux.jl)
    4. Conformal Prediction for Time Series
  3. Questions

🤖 Conformal Chatbot

🔎 Conformal Counterfactual Explanations

🌐 Conformal Bayesian Neural Networks

📈 Conformal Time Series

🐶 Taija

Methodologies and open-source tools that help researchers and practitioners assess the trustworthiness of predictive models.

Towards Trustworthy AI in Julia
  1. CounterfactualExplanations.jl (JuliaCon 2022)
  2. ConformalPrediction.jl (JuliaCon 2023 — I hope!)
  3. LaplaceRedudx.jl (JuliaCon 2022)
  4. AlgorithmicRecourseDynamics.jl

… contributions welcome! 😊

📚 More Reading

Image Sources

  • Copyright for stock images belongs to TU Delft.
  • All other images, graphics or animations were created by us.

References